← Back to HEINI

HEINI Enterprise · AI server for DACH

Your AI. Your server. Your decision.

Run your own language and generative AI models on a GPU server at your premises, connected to HEINI.

Book a GPU server consultation See how it connects ↓

Your hardware. A clear operating budget.

HEINI connects your own models with the work in your business. Choose the right starting scope.

Your own AI server

Company

2 GPUs · one legal entity

€216,500 upfront, net

HEINI software from €12,900 / month

from €154,800 / year in advance

Company group

4 GPUs · up to 5 agreed legal entities

€517,000 upfront, net

HEINI software from €24,900 / month

from €298,800 / year in advance

Setup and operation are additional. Updates, agreed support and users included.

5-year manufacturer warranty. HEINI exclusively distributes this server offer in Germany, Austria and Switzerland.

Indicative hardware prices, net plus applicable VAT. Configuration and delivery scope as agreed in your HEINI offer.

Local model calls: no external LLM token fees. HEINI software, electricity, operation and any external services remain separate costs.

Calculate your own costs

What does your own AI operation cost?

Buy upfront. Operate monthly. See both together here.

Calculated only in your browser. Your figures are not sent or stored.

Adjust the calculation

Changing the system resets your assumptions and comparison values.

Example over 5 years · Company · 2 GPUs

HEINI software from €154,800 per year, paid annually in advance. Recurring costs shown as monthly equivalents. All amounts net, plus applicable VAT.

Upfront
€241,500
Server + setup
Recurring · monthly equivalent
€14,400
Software + operation + external services
Total over 5 years
€1,105,500
Upfront and ongoing costs

An average of €18,425 per month. Not a lease payment or a total monthly bill.

Adjust assumptions and comparison
Adjust assumptions

Assumption, not a fixed fee. Include migration and special services here.

Electricity, cooling, space, administration and backup. Do not count included software support twice.

Assumption at €0: all model calls considered run locally. Otherwise add remaining costs.

Calculation horizon, not a new contract term.

Compare with an alternative

Compare the same scope: software, AI usage and operation together, all net. Treat any ERP and services you still need consistently in both options. Do not add included AI allowances twice.

Leave the monthly comparison blank to show only Private costs. Clear the field to remove the comparison.

Hardware · indicative, upfront
€216,500
Setup · assumption, upfront
€25,000
HEINI software · from, per month
€12,900
Infrastructure · assumption, per month
€1,500
External services · assumption, per month
€0

At €0 for external services: all model calls considered run locally. Otherwise add costs under “Adjust assumptions”. Operation includes electricity, cooling, space, infrastructure administration and backup.

Indicative hardware prices, net plus applicable VAT. Configuration and delivery scope as agreed in your HEINI offer. HEINI software is paid annually in advance. The horizon is not a contract term; this is not a monthly payment schedule. Constant prices, without financing, discounting, tax effects or residual value. Setup is an adjustable project budget, not a fixed fee.

Discuss your server use case →

Only the selected system size becomes the consultation topic. Your figures stay here.

Explained operating model · scope by agreement

Your compute. A clear path for your data.

Example conversation

  1. You · at work

    Can our own language and generative AI models run on our premises?

  2. HEINI

    For enterprise customers we offer a dedicated GPU server, exclusively in DACH. Models, data paths and operations are agreed. Local inference does not mean the entire HEINI platform runs locally.

The prepared work

HEINI connects the agreed workflow with your AI server. Inference for the configured models runs there.

Your next decision

You define models, data paths, access and approvals. The actual configuration determines which data crosses each system boundary.

The benefit for your team

Less searching. More clarity.

  1. 01

    Models matched to the task

    Your own language and GenAI models for agreed applications. Model size, licence and quality are checked before use.

  2. 02

    Hardware at your premises

    Your compute can be sized for usage, concurrency and required memory.

  3. 03

    HEINI as your point of contact

    Consultation, connection and agreed support form one coordinated plan. Your team keeps control.

A practical start

A clear path into your business.

HEINI distributes this server offer exclusively in Germany, Austria and Switzerland. Hardware, model operation and service are tailored to your requirements.

  1. 01

    Discuss requirements

    Tasks, models, data types and concurrent users define the requirements.

  2. 02

    Size the server

    GPU memory, compute, cooling, power and networking are planned to fit.

  3. 03

    Define operations

    Access, data paths, updates, backups, support and failure behaviour are agreed.

  4. 04

    Accept the connection

    We assess model responses and the HEINI workflow against agreed cases before your team takes over.

Before you decide

Good questions. Clear answers.

01Does this run all of HEINI on my premises?

The GPU server initially covers the agreed model operation. It does not imply a fully local deployment of HEINI. Standard hosting and connected services are identified separately in the operating plan.

02Is all data guaranteed to stay in-house?

That depends on the agreed configuration. Data paths, telemetry, external services and failure behaviour must be explicitly defined and tested. Owning a server alone does not guarantee isolated operation.

03Which models and GPUs are included?

The proposal is tailored to your requirements. Model licences, quality, memory and throughput inform the selection. Supported text, image and other GenAI applications are explicitly agreed.

04What does a private HEINI AI server cost?

Company: 2 GPUs, €216,500 hardware upfront and HEINI software from €12,900 net monthly equivalent (€154,800 annually in advance) for one legal entity. Company group: 4 GPUs, €517,000 hardware upfront and software from €24,900 net monthly equivalent (€298,800 annually in advance) for up to 5 agreed legal entities. Setup, infrastructure and external services are additional. Indicative hardware prices, net plus applicable VAT. Configuration and delivery scope as agreed in your HEINI offer.

05What is included in the Private offer?

5-year manufacturer warranty. HEINI exclusively distributes this server offer in Germany, Austria and Switzerland. HEINI software is paid annually in advance as a subscription with updates and agreed support, without per-seat pricing. Local model calls incur no external LLM token fees. Remaining external services and the scope of a local platform are agreed separately.

06Where is this offer available?

HEINI distributes this server offer exclusively in DACH: Germany, Austria and Switzerland. Delivery, installation and service scope are stated in your individual proposal.

Bring your use case.

In 30 minutes we clarify the goal, required connections and a useful first step. No sensitive production data is needed.

Book a GPU server consultation